Fast & Curious
Gilles Debunne

I almost missed it

My deadline.

Monday was a holiday, but today was supposed to be my first post.

Indeed, the idea is to use this blog to push me to discover and then write about things I've learned and experienced.

Procrastination was doing a good job, it is now 10PM. But it worked: the totally virtual peer pressure pushed me to write these lines.

Being overly optimistic, I'll try to post every week day. We'll see how it goes. That's about 200 posts in a year. If I learn half as many tricks, I'll be a better programmer.

It's already working

Since I started this blog, I've been playing with Jekyll, the Markdown syntax and GitHub user pages, among other things (such as forcing myself to write in English).

These are fantastic tools, and I'm glad I have added these to my tool belt.

Let's see what we'll talk about tomorrow.

I want to be a better programer

We're Saturday morning, 8:43AM, and I'm typing this after yet an other (too) delicious breakfast at my airbnb place.

I want to be a better programmer.

That's what I had in my head when I woke up this morning. Devoxx ended yesterday and I didn't enjoy it that much. But as I said in my conclusion, the benefit of this conference is that it motivated me, pushed me to my keyboard. I realize the reason is that I want to be like these guys on stage. I want to use and talk about these new cool technologies, I want to be more efficient at what I do. (I want to be famous and recognized in the coridors of the conference?).

I need to train and practice.

Since this idea came to mind, thousands of others joined in. I've created a movie, imagining some scenes about what is supposed to happen next, what I'll do and say, how I'll handle this project. That's what happens to me when I start thinking about something new and fun, and it is usually a good sign. But it is a long journey and we'll see, it may as well stop in 5 minutes, so let's be lean.

I wrote it down, and no one will read this, unless I actually give this project a chance. We'll see what happens on Monday. And in the meantime, let's go and enjoy this last day in Paris.

À nous deux, maintenant !

Devoxx 2014 - Day 3

Keynotes

An inspiring first opening keynote from a musician who's also a programer. Pushing us to be more creative in our work, pursuing our own personal goal.

Conferences

Apps for kids

A list of good practices when designing apps for little kids. Mostly common sense, but worth getting the slides to use as a checklist.

Parsers

Although we believe we have learned them at school, parsers are not that simple. A dynamic presentation proved it, going from LL1 to PEG and finally to GLL (Generalized LL) to handle the ambiguities of C++.

Quickies

Git deliver

A plugin to use git as a delivery tool, like on heroku. Git with hook scripts, mostly for source and dynamic language deployment.

Maths for recommandations

Weighted averages, similarities between users and between products, leading to recommandations. The maths are pretty simple indeed, but more heuristics are needed for a real applications based on domain specificities.

Flexible scope

Probably the presentation I've preferred in the conference. Check it out on Parley's when it's available.

Agile needs to spread to business, so that the whole industry process can be tuned, and not just the software process. This reactive capacity is needed since Time / Humans / Locale are unpredictable dimensions and must be coped by flexibility.

We are mostly using negative metrics in our industry, such as # of bugs, duplicated code or even velocity. But they are only alerts that something is wrong, not a proof that we are right, and only trying to optimize these is bad. He makes an analogy with blood pressure for humans.

We need higher picture victory conditions, such as a change condition in a user story to show its value and decide on its success. An example of flexibility is being able to remove user stories, features and abandon code based on feedback.

User stories are survivable experiments.

One needs a global picture to drive flexibility. Analogy with a GPS, targeting a goal, but able to able to adapt live to the situation.

Nice stories, references and talk. I've noted "Adapt" from Tim Hartford.

33 Tricks

After a long introduction, he presented a bunch of inspiring and useful libraries and tricks (Lumbock, Guava, LambdaJ,...), to freshen our coding habits.

Go for Javaïstes

Really cool presentation, tailored for Java programers and presenting the Go language. Detailing the strong choices that were made and showing some sample code detailing the core system of concurrency based on message passing that act as locks and make concurrent code seemingly easy.

Extremist programming

How 'weird' people actually change the world. New disruptive ideas emerge out of some chaos and we should try them out instead of repeating the same practices over and over. Like trying out a new crazy languages, adding constraints or looking at other fields. Follow their @weirddev for such a daily kata.

Les Cast coders

Amazing how these guys on stage have exactly the same voice as those on one of my podcasts.

Conclusion

Technology is at the core of the digital revolution, but the programer (us), does not seems to be at its center.

He is a mere executant, not well understood by the other actors, who don't understand this programming activity either.

There is a communication issue here, between the idea and the code, which prevents new projects to bloom. Humans are part of that chain, and although there are less and less different roles involved, the programer is an essential link here. His success and the success of his projects will depend on his ability to communicate with the other actors.

Personal

I'm not fitted for this kind of conferences. The pace and format of sessions makes me want to watch them on Parleys, or better, get the slides with speaker notes.

Still, strangely, watching other developers talk about their work has the effect of pushing me to my keyboard, to do my own programs, try some of the presented techniques, prove that I'm too a programer.